def run_gc_depth(genome, fastq_list, name, window, thread, job_type,
concurrent, refresh, work_dir, out_dir):
genome, fastq_list = check_paths([genome, fastq_list])
sort_bam, genome = bwa_mem(fastq_list=fastq_list,
genome=genome,
name=name,
number=5000000,
data_type='',
thread=thread,
job_type=job_type,
concurrent=concurrent,
refresh=refresh,
work_dir=work_dir,
out_dir=work_dir)
sort_bam = check_paths(sort_bam)
dag = DAG("gc_depth")
gc_depth_task, gc_depth_png = stat_gc_depth_task(genome=genome,
bam=sort_bam,
name=name,
window=window,
job_type=job_type,
work_dir=work_dir,
out_dir=out_dir)
dag.add_task(gc_depth_task)
do_dag(dag, concurrent, refresh)
return gc_depth_png
def stat_gc_depth_task(genome, bam, name, window, job_type, work_dir, out_dir):
bam = check_paths(bam)
task = Task(id="stat_coverage",
work_dir=work_dir,
type=job_type,
option="-pe smp 1",
script="""
export PATH={samtools}:{python}:$PATH
samtools depth -aa {bam} > {work_dir}/{name}.depth
python {script}/stat_coverage.py -i {work_dir}/{name}.depth -d 1,5,10,20 -o {out_dir}/{name}.coverage.xlsx
python {script}/stat_length_gc.py -d {work_dir}/{name}.depth -g {genome} -n {out_dir}/{name}
python {script}/stat_gc_depth.py -d {work_dir}/{name}.depth -g {genome} -b 1000 -w 5000 -e 100 -n {work_dir}/{name}
python {script}/draw_depth_gc.py -gcd {work_dir}/{name}.stat_gc_depth.tsv -n {out_dir}/{name}
#python {script}/plot_gc_depth.py -gcd {work_dir}/{name}.stat_gc_depth.tsv -n {out_dir}/{name}
""".format(samtools=SAMTOOLS_BIN,
script=SCRIPTS,
python=PYTHON_BIN,
genome=genome,
bam=bam,
name=name,
window=window,
work_dir=work_dir,
out_dir=out_dir))
return task, os.path.join(out_dir, "%s.gc_depth.png" % name)
def minimap(r1, r2, genome, name, split, platform, number, thread, job_type,
concurrent, refresh, work_dir, out_dir):
genome = check_path(genome)
work_dir = mkdir(work_dir)
out_dir = mkdir(out_dir)
r1 = check_paths(r1)
if r2 != "":
r2 = check_paths(r2)
options = {"software": OrderedDict(), "database": OrderedDict()}
data_work = mkdir(os.path.join(work_dir, '00_data'))
reads = split_data(r1=r1,
r2=r2,
name=name,
number=number,
job_type=job_type,
work_dir=data_work,
out_dir=out_dir)
bam, option = run_minimap(reads=reads,
genome=genome,
platform=platform,
name=name,
split=split,
thread=thread,
job_type=job_type,
concurrent=concurrent,
refresh=refresh,
work_dir=work_dir,
out_dir=out_dir)
options["software"] = option
with open(os.path.join(out_dir, "minimap2.json"), "w") as fh:
json.dump(options, fh, indent=2)
return bam
def run_gc_depth(genome, r1, r2, name, platform, split, window, thread,
job_type, concurrent, refresh, work_dir, out_dir):
genome = check_path(genome)
r1 = check_paths(r1)
r2 = check_paths(r2)
sort_bam = minimap(r1=r1,
r2=r2,
genome=genome,
name=name,
split=split,
platform=platform,
number=5000000,
thread=thread,
job_type=job_type,
concurrent=concurrent,
refresh=refresh,
work_dir=work_dir,
out_dir=out_dir)
sort_bam = check_paths(sort_bam)
dag = DAG("gc_depth")
gc_depth_task, gc_depth_png = stat_gc_depth_task(genome=genome,
bam=sort_bam,
name=name,
window=window,
job_type=job_type,
work_dir=work_dir,
out_dir=out_dir)
dag.add_task(gc_depth_task)
do_dag(dag, concurrent, refresh)
return gc_depth_png
def bwa_mem(fastq_list, genome, name, number, data_type, thread, job_type,
concurrent, refresh, work_dir, out_dir):
genome, fastq_list = check_paths([genome, fastq_list])
work_dir = mkdir(work_dir)
out_dir = mkdir(out_dir)
dag = DAG("split_ngs")
split_work = mkdir(os.path.join(work_dir, "00_data"))
split_out = mkdir(os.path.join(out_dir, "00_data"))
splitfp_task, fq_path, r1_name, r2_name = split_ngs_task(
fastq_list=fastq_list,
name=name,
number=number,
data_type=data_type,
job_type=job_type,
work_dir=split_work,
out_dir=split_out)
dag.add_task(splitfp_task)
do_dag(dag, concurrent, refresh)
dag = DAG("bwa_mem")
index_task, bwa_tasks, merge_task, sorted_bam, genome = run_bwa_mem(
fq_path=fq_path,
r1_name=r1_name,
r2_name=r2_name,
genome=genome,
name=name,
thread=thread,
job_type=job_type,
work_dir=work_dir,
out_dir=out_dir)
dag.add_task(index_task)
dag.add_task(*bwa_tasks)
dag.add_task(merge_task)
index_task.set_downstream(*bwa_tasks)
merge_task.set_upstream(*bwa_tasks)
do_dag(dag, concurrent, refresh)
return sorted_bam, genome
def run_survey(r1, r2, name, trim, kingdom, kmer_length, sample_depth, thread,
asm, window, job_type, queue, concurrent, refresh, work_dir,
out_dir):
work_dir = mkdir(work_dir)
out_dir = mkdir(out_dir)
r1 = check_paths(r1)
r2 = check_paths(r2)
dag = DAG("survey_qc")
merge_task, qc_task, cont_task, result_task, clean1, clean2, quality, content, gc, stat_qc, poll_png, poll_tsv = ngs_qc_tasks(
name=name,
r1=r1,
r2=r2,
trim=trim,
thread=thread,
job_type=job_type,
work_dir=work_dir,
out_dir=out_dir)
data_work = mkdir(os.path.join(work_dir, "choose_data"))
freq_task1, histo1, kmer_stat, estimate1 = kmerfreq_task(
r1=clean1,
r2=clean2,
name=name,
kmer_length=kmer_length,
thread=thread,
job_type=job_type,
work_dir=data_work,
out_dir=data_work)
dag.add_task(merge_task)
dag.add_task(qc_task)
qc_task.set_upstream(merge_task)
dag.add_task(cont_task)
dag.add_task(result_task)
dag.add_task(freq_task1)
freq_task1.set_upstream(qc_task)
cont_task.set_upstream(qc_task)
result_task.set_upstream(qc_task)
do_dag(dag, concurrent, refresh)
for line in read_tsv(kmer_stat):
if line[0] == "kmer_depth":
kmer_depth = int(line[1])
if sample_depth > kmer_depth:
LOG.debug(
'The amount of sequencing data may be insufficient. Sequencing depth is only %s X'
% kmer_depth)
sample_depth = kmer_depth
proportion = sample_depth * 1.0 / kmer_depth
dag = DAG("survey")
choose_task, freq_task, heter_task, jellyfish_task, gse_scope_task, denovo_task, stat_heter, heter_png, scope_txt, gse_txt, scope_png, gse_png, stat_genome, genome, ngs_list = kmer_denovo_tasks(
r1=clean1,
r2=clean2,
name=name,
kmer_length=kmer_length,
proportion=proportion,
kingdom=kingdom,
thread=thread,
job_type=job_type,
queue=queue,
work_dir=work_dir,
out_dir=out_dir)
if asm == "true":
dag.add_task(denovo_task)
else:
genome = "false"
stat_genome = "false"
ngs_list = "false"
dag.add_task(choose_task)
dag.add_task(freq_task)
dag.add_task(heter_task)
freq_task.set_upstream(choose_task)
dag.add_task(jellyfish_task)
jellyfish_task.set_upstream(choose_task)
dag.add_task(gse_scope_task)
heter_task.set_upstream(freq_task)
gse_scope_task.set_upstream(jellyfish_task)
do_dag(dag, concurrent, refresh)
if ngs_list == "false":
print("Genomics are not assembled")
gc_depth_png = heter_png
else:
depth_work = mkdir(os.path.join(work_dir, "05_GC-depth"))
depth_out = mkdir(os.path.join(out_dir, "05_GC-depth"))
gc_depth_png = run_gc_depth(genome=genome,
fastq_list=ngs_list,
name=name,
window=window,
thread=thread,
job_type=job_type,
concurrent=concurrent,
refresh=refresh,
work_dir=depth_work,
out_dir=depth_out)
run_report(name, asm, kmer_length, stat_qc, quality, content, gc, poll_tsv,
poll_png, stat_heter, heter_png, scope_txt, gse_txt, scope_png,
gse_png, stat_genome, gc_depth_png, out_dir)
return stat_qc, quality, content, gc, poll_png, poll_tsv, stat_heter, heter_png, scope_txt, gse_txt, scope_png, gse_png, stat_genome
def run_survey(r1,
r2,
name,
trim,
kingdom,
mode,
cratio,
kmer_length,
kmer_depth,
thread,
asm,
window,
job_type,
queue,
concurrent,
refresh,
work_dir,
out_dir,
split=""):
work_dir = mkdir(work_dir)
out_dir = mkdir(out_dir)
r1 = check_paths(r1)
r2 = check_paths(r2)
clean1, clean2, taxid, stat_qc, quality, content, gc, cont_tsv, cont_png = run_ngs_qc(
r1=r1,
r2=r2,
name=name,
trim=trim,
kingdom=kingdom,
thread=thread,
job_type=job_type,
concurrent=concurrent,
refresh=refresh,
work_dir=os.path.join(work_dir, "01_data"),
out_dir=os.path.join(out_dir, "01_data"))
stat_heter, heter_png, scope_txt, gse_txt, scope_png, gse_png, stat_genome, gc_depth = run_kmer_denovo(
r1=[clean1],
r2=[clean2],
taxid=taxid,
name=name,
mode=mode,
cratio=cratio,
kmer_length=kmer_length,
kmer_depth=kmer_depth,
kingdom=kingdom,
asm=asm,
window=window,
thread=thread,
job_type=job_type,
queue=queue,
concurrent=concurrent,
refresh=refresh,
work_dir=work_dir,
out_dir=out_dir,
split=split,
platform="illumina")
run_report(name, asm, kmer_length, stat_qc, quality, content, gc, cont_tsv,
cont_png, stat_heter, heter_png, scope_txt, gse_txt, scope_png,
gse_png, stat_genome, gc_depth, out_dir)
def run_filter_contamination(r1, r2, name, kmer_length, kmer_depth, taxid, kingdom, thread, job_type, concurrent, refresh, work_dir, out_dir, split, mode="fast", cratio=10):
work_dir = mkdir(work_dir)
out_dir = mkdir(out_dir)
r1 = check_paths(r1)
r2 = check_paths(r2)
taxid = check_path(taxid)
options = {
"software": OrderedDict(),
"database": OrderedDict()
}
option, r1, r2 = choose_data(
r1=r1,
r2=r2,
name=name,
kmer_length=kmer_length,
kmer_depth=kmer_depth,
thread=thread,
job_type=job_type,
concurrent=concurrent,
refresh=refresh,
work_dir=work_dir,
out_dir=out_dir)
options["software"].update(option)
if mode!="fast":
work_dict = {
"data": "00_data",
"ref": "01_ref",
"ump": "02_ump"
}
for k, v in work_dict.items():
mkdir(os.path.join(work_dir, v))
reads = split_data(
r1=r1,
r2=r2,
name=name,
number=2000000,
job_type=job_type,
work_dir=os.path.join(work_dir, work_dict["data"]),
concurrent=concurrent,
refresh=refresh,
out_dir=out_dir,
platform="illumina")
dag = DAG("unmap_data")
ref_task, ref= obtain_contamination_task(
taxid=taxid,
name=name,
kingdom=kingdom,
job_type=job_type,
work_dir=os.path.join(work_dir, work_dict["ref"]),
out_dir=out_dir,
mode=mode,
cratio=cratio)
dag.add_task(ref_task)
unmap_tasks, reads, option = create_unmap_tasks(
name=name,
reference=ref,
reads=reads,
thread=thread,
job_type=job_type,
work_dir=os.path.join(work_dir, work_dict["ump"]),
out_dir=out_dir,
split=split)
dag.add_task(*unmap_tasks)
ref_task.set_downstream(*unmap_tasks)
do_dag(dag, concurrent, refresh)
options["software"].update(option)
reads = [reads]
else:
reads = [r1, r2]
return reads, options
def run_kmer_denovo(r1, r2, name, kingdom, kmer_length, sample_depth, thread,
asm, window, job_type, queue, concurrent, refresh,
work_dir, out_dir):
work_dir = mkdir(work_dir)
out_dir = mkdir(out_dir)
r1 = check_paths(r1)
r2 = check_paths(r2)
if r1[0].endswith(".gz") or r2[0].endswith(".gz"):
tools = "zcat"
else:
tools = "cat"
dag_data = DAG("survey_data")
data_work = mkdir(os.path.join(work_dir, "choose_data"))
cat_data_task, clean1, clean2 = merge_raw_data_task(name=name,
r1=" ".join(r1),
r2=" ".join(r2),
tools=tools,
job_type=job_type,
work_dir=data_work,
out_dir=data_work)
freq_task1, histo1, kmer_stat, estimate1 = kmerfreq_task(
r1=clean1,
r2=clean2,
name=name,
kmer_length=17,
thread=thread,
job_type=job_type,
work_dir=data_work,
out_dir=data_work)
dag_data.add_task(cat_data_task)
dag_data.add_task(freq_task1)
freq_task1.set_upstream(cat_data_task)
do_dag(dag_data, concurrent, refresh)
for line in read_tsv(kmer_stat):
if line[0] == "kmer_depth":
kmer_depth = int(line[1])
if sample_depth > kmer_depth:
LOG.debug(
'The amount of sequencing data may be insufficient. Sequencing depth is only %s X'
% kmer_depth)
sample_depth = kmer_depth
proportion = sample_depth * 1.0 / kmer_depth
dag = DAG("survey")
choose_task, freq_task, heter_task, jellyfish_task, gse_scope_task, denovo_task, stat_heter, heter_png, scope_txt, gse_txt, scope_png, gse_png, stat_genome, genome, ngs_list = kmer_denovo_tasks(
r1=clean1,
r2=clean2,
name=name,
kmer_length=kmer_length,
proportion=proportion,
kingdom=kingdom,
thread=thread,
job_type=job_type,
queue=queue,
work_dir=work_dir,
out_dir=out_dir)
if asm == "true":
dag.add_task(denovo_task)
else:
genome = "false"
stat_genome = "false"
ngs_list = "false"
dag.add_task(choose_task)
dag.add_task(freq_task)
dag.add_task(heter_task)
freq_task.set_upstream(choose_task)
dag.add_task(jellyfish_task)
jellyfish_task.set_upstream(choose_task)
dag.add_task(gse_scope_task)
heter_task.set_upstream(freq_task)
gse_scope_task.set_upstream(jellyfish_task)
do_dag(dag, concurrent, refresh)
if ngs_list == "false":
print("Genomics are not assembled")
gc_depth_png = heter_png
else:
depth_work = mkdir(os.path.join(work_dir, "05_GC-depth"))
depth_out = mkdir(os.path.join(out_dir, "05_GC-depth"))
gc_depth_png = run_gc_depth(genome=genome,
fastq_list=ngs_list,
name=name,
window=window,
thread=thread,
job_type=job_type,
concurrent=concurrent,
refresh=refresh,
work_dir=depth_work,
out_dir=depth_out)
return stat_heter, heter_png, scope_txt, gse_txt, scope_png, gse_png, stat_genome
def kmer_denovo_tasks(r1, r2, name, kmer_length, proportion, kingdom, thread,
job_type, queue, work_dir, out_dir):
work_dir = mkdir(work_dir)
out_dir = mkdir(out_dir)
clean1 = check_paths(r1)
clean2 = check_paths(r2)
work_dict = {
"choose": "choose_data",
"gse_scope": "02_gse_scope",
"kmerfreq": "03_Kmerfreq",
"denovo": "04_Soapdenovo",
"gc_depth": "05_GC-depth"
}
choose_work = mkdir(os.path.join(work_dir, work_dict["choose"]))
choose_task, choose_r1, choose_r2, choose_r = sample_fastq_task(
r1=clean1,
r2=clean2,
proportion=proportion,
name=name,
job_type=job_type,
work_dir=choose_work)
heter_work = mkdir(os.path.join(work_dir, work_dict["kmerfreq"]))
heter_out = mkdir(os.path.join(out_dir, work_dict["kmerfreq"]))
freq_task, histo, kmer_depth, estimate = kmerfreq_task(r1=choose_r1,
r2=choose_r2,
name=name,
kmer_length=17,
thread=thread,
job_type=job_type,
work_dir=heter_work,
out_dir=heter_out)
heter_task, stat_heter, heter_png = get_heterozygosity_task(
histo=histo,
estimate=estimate,
kingdom=kingdom,
name=name,
job_type=job_type,
work_dir=heter_work,
out_dir=heter_out)
scope_work = mkdir(os.path.join(work_dir, work_dict["gse_scope"]))
scope_out = mkdir(os.path.join(out_dir, work_dict["gse_scope"]))
jellyfish_task, histogram = get_jellyfish_task(fastq=choose_r,
name=name,
depth=40 * 100,
thread=thread,
job_type=job_type,
work_dir=scope_work,
out_dir=scope_out)
gse_scope_task, scope_txt, gse_txt, scope_png, gse_png = get_gse_scope_task(
histogram=histogram,
name=name,
kmer_length=kmer_length,
job_type=job_type,
work_dir=scope_work,
out_dir=scope_out)
denovo_work = mkdir(os.path.join(work_dir, work_dict["denovo"]))
denovo_out = mkdir(os.path.join(out_dir, work_dict["denovo"]))
denovo_task, genome, stat_genome, ngs_list = soapdenovo_task(
r1=clean1,
r2=clean2,
name=name,
thread=thread * 2,
queue=queue,
job_type=job_type,
work_dir=denovo_work,
out_dir=denovo_out)
return choose_task, freq_task, heter_task, jellyfish_task, gse_scope_task, denovo_task, stat_heter, heter_png, scope_txt, gse_txt, scope_png, gse_png, stat_genome, genome, ngs_list
def run_kmer_denovo(r1,
r2,
taxid,
name,
mode,
cratio,
kmer_length,
kmer_depth,
kingdom,
asm,
window,
thread,
job_type,
queue,
concurrent,
refresh,
work_dir,
out_dir,
split,
platform="illumina"):
work_dir = mkdir(work_dir)
out_dir = mkdir(out_dir)
r1 = check_paths(r1)
r2 = check_paths(r2)
work_dict = {
"contamination": "01_contamination",
"gse_scope": "02_gse_scope",
"kmerfreq": "03_Kmerfreq",
"denovo": "04_Soapdenovo",
"gc_depth": "05_GC-depth"
}
for k, v in work_dict.items():
mkdir(os.path.join(work_dir, v))
if k == "contamination":
continue
mkdir(os.path.join(out_dir, v))
reads, options = run_filter_contamination(r1=r1,
r2=r2,
name=name,
kmer_length=kmer_length,
kmer_depth=kmer_depth,
taxid=taxid,
kingdom=kingdom,
thread=thread,
job_type=job_type,
concurrent=concurrent,
refresh=refresh,
work_dir=os.path.join(
work_dir,
work_dict["contamination"]),
out_dir=out_dir,
mode=mode,
cratio=cratio,
split=split)
dag = DAG("kmer_denovo")
jellyfish_task, gse_scope_task, scope_txt, gse_txt, scope_png, gse_png, option = gse_scope(
reads=" ".join(reads),
name=name,
kmer_length=kmer_length,
thread=thread,
job_type=job_type,
work_dir=os.path.join(work_dir, work_dict["gse_scope"]),
out_dir=os.path.join(out_dir, work_dict["gse_scope"]),
mode=mode)
options["software"].update(option)
dag.add_task(jellyfish_task)
dag.add_task(gse_scope_task)
kmerfreq_task, heter_task, stat_heter, heter_png, option = kmerfreq(
reads=" ".join(reads),
name=name,
kingdom=kingdom,
kmer_length=kmer_length,
thread=thread,
job_type=job_type,
work_dir=os.path.join(work_dir, work_dict["kmerfreq"]),
out_dir=os.path.join(out_dir, work_dict["kmerfreq"]))
options["software"].update(option)
dag.add_task(kmerfreq_task)
dag.add_task(heter_task)
denovo_task, genome, stat_genome, option = create_soapdenovo_task(
r1=" ".join(r1),
r2=" ".join(r2),
name=name,
thread=thread,
queue=queue,
job_type=job_type,
work_dir=os.path.join(work_dir, work_dict["denovo"]),
out_dir=os.path.join(out_dir, work_dict["denovo"]))
if asm == "true":
dag.add_task(denovo_task)
else:
genome = "false"
stat_genome = "false"
do_dag(dag, concurrent, refresh)
if asm == "true":
gc_depth = run_gc_depth(genome=genome,
r1=" ".join(r1),
r2=" ".join(r2),
name=name,
platform=platform,
split="no_split",
window=window,
thread=thread,
job_type=job_type,
concurrent=concurrent,
refresh=refresh,
work_dir=os.path.join(work_dir,
work_dict["gc_depth"]),
out_dir=os.path.join(out_dir,
work_dict["gc_depth"]))
else:
gc_depth = heter_png
with open(os.path.join(out_dir, "kmer_denovo.json"), "w") as fh:
json.dump(options, fh, indent=2)
return stat_heter, heter_png, scope_txt, gse_txt, scope_png, gse_png, stat_genome, gc_depth